On Thu, 04 Apr 2013 11:49:59 +0100 Tom Hacohen <[email protected]> said:

> On 04/04/13 11:43, Carsten Haitzler (The Rasterman) wrote:
> > On Thu, 04 Apr 2013 08:39:24 +0100 Tom Hacohen <[email protected]>
> > said:
> >
> >> On 04/04/13 00:52, Carsten Haitzler (The Rasterman) wrote:
> >>> On Wed, 03 Apr 2013 17:26:42 +0100 Tom Hacohen <[email protected]>
> >>> said:
> >>>
> >>>> On 28/03/13 10:49, Carsten Haitzler (The Rasterman) wrote:
> >>>>> On Thu, 28 Mar 2013 09:56:40 +0000 Michael Blumenkrantz
> >>>>> <[email protected]> said:
> >>>>>
> >>>>> thats cool. i just had to be grumpy about not having a bug report that
> >>>>> told me what to look at instantly. i have found another bug. single
> >>>>> letter words dont find word end markers.
> >>>>
> >>>> I just checked it, and it works for me:
> >>>> #include <stdlib.h>
> >>>> #include <wchar.h>
> >>>> #include <stdio.h>
> >>>> #include <wordbreak.h>
> >>>>
> >>>> int main()
> >>>> {
> >>>>         {
> >>>>            const char *lang = "";
> >>>>            wchar_t *text = L"This is a test";
> >>>>            size_t len = wcslen(text);
> >>>>            char *breaks = malloc(len);
> >>>>            size_t i;
> >>>>
> >>>>            printf("%ls\n", text);
> >>>>
> >>>>            set_wordbreaks_utf32((const utf32_t *) text, len, lang,
> >>>> breaks); for (i = 0 ; i < len ; i++)
> >>>>               printf("%d", (int) breaks[i]);
> >>>>            printf("\n");
> >>>>         }
> >>>>       return 0;
> >>>> }
> >>>>
> >>>> The output is:
> >>>> This is a test
> >>>> 11100100001110
> >>>>
> >>>> 1s meaning no break, 0s meaning break here. It does break correctly
> >>>> around the "a". Could you elaborate more on the bug you were seeing?
> >>>
> >>> no NON-breaks around "a". you can't tell that there is a word there at
> >>> all. it may as well be "   " (all spaces). :)
> >>>
> >>>> Cheers,
> >>>> Tom.
> >>>>
> >>>
> >>>
> >>
> >> Yeah, well, you know it using other means. Unfortunately it's beyond the
> >> scope of the word breaking algorithm... There are no word breaks there,
> >> thus the algorithm produces none. You probably need to just skip whites
> >> in your code, not only rely on the wordbreak data when "merging" the
> >> whites.
> >
> > and that is a problem as the word next/prev stuff relies on this.. and what
> > are "whites" then? (from the word breaking point of view)... eg ' is NOT
> > white. ( is. etc....
> >
> >
> 
> Well, word breaking has nothing to do with whites (well, they happen to 
> be in a class that separates words, but that's it). I wouldn't change 
> the word next/prev functions themselves, I'd just change the way they 
> are used in edje/elm. I.e, something like:
> 
> If (is_white(cur_char))
> {
>       skip_whites;
>       skip_word;
> }
> else
> {
>       skip_word;
>       skip_whites;
> }     

and therein lies the rub. "what is a white" when it comes to word separattion
assuming white == separator = eg " ", "/t", "\n", ")", "(", "." etc.

> ------------------------------------------------------------------------------
> Minimize network downtime and maximize team effectiveness.
> Reduce network management and security costs.Learn how to hire 
> the most talented Cisco Certified professionals. Visit the 
> Employer Resources Portal
> http://www.cisco.com/web/learning/employer_resources/index.html
> _______________________________________________
> enlightenment-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> 


-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    [email protected]


------------------------------------------------------------------------------
Minimize network downtime and maximize team effectiveness.
Reduce network management and security costs.Learn how to hire 
the most talented Cisco Certified professionals. Visit the 
Employer Resources Portal
http://www.cisco.com/web/learning/employer_resources/index.html
_______________________________________________
enlightenment-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Reply via email to